Pith. sign in

REVIEW 4 minor 68 references

Non-Gaussian inputs push bosonic quantum communication past the thermal threshold where every single-mode Gaussian state fails.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

T0 review · grok-4.5

2026-07-31 00:32 UTC pith:J6CBP2Q2

load-bearing objection They close the 1999 Gaussian-optimality gap for thermal-attenuator coherent information and give the first certified non-Gaussian witnesses with Q>0 between antidegradability and the thermal threshold.

arxiv 2607.27449 v1 pith:J6CBP2Q2 submitted 2026-07-29 quant-ph

Bosonic quantum communication beyond the thermal threshold

Francesco Anna Mele , Giuseppe Catalano , Marco Fanizza , Vittorio Giovannetti , Ludovico Lami This is my paper
classification quant-ph
keywords quantum capacitythermal attenuatorcoherent informationnon-Gaussian statesbosonic Gaussian channelsantidegradabilityHolevo-Werner bound
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

For decades the best known lower bound on the quantum capacity of the thermal attenuator came from thermal input states, and that bound vanishes below a thermal threshold. This paper first proves that no single-mode Gaussian state can beat that bound, then exhibits an explicit rank-two non-Gaussian state, supported on only six Fock levels, whose coherent information is strictly positive at the threshold point where every Gaussian input yields zero. Because a positive coherent information is an achievable rate, the quantum capacity itself is positive in a previously unsettled high-noise window between the antidegradability line and the thermal threshold. At one thermal photon the authors certify positivity down to transmissivity 0.7841, while the channel is forced to zero capacity only at or below 0.75. The result shows that a complete theory of noisy bosonic quantum communication cannot stay inside the Gaussian world.

Core claim

The long-standing Holevo–Werner thermal-state lower bound is exactly the supremum of the coherent information over all single-mode Gaussian inputs; yet a concrete non-Gaussian state achieves a strictly larger value, so the quantum capacity of the thermal attenuator is positive in a parameter region where that Gaussian optimum vanishes.

What carries the argument

An explicit rank-two non-Gaussian mixture on six Fock levels, combined with a cutoff-erasure channel that reduces the coherent-information calculation to finite matrices whose positivity is certified by interval arithmetic and Gershgorin eigenvalue enclosures.

Load-bearing premise

The strict positivity proof rests on a computer-assisted interval-arithmetic certificate of two finite-dimensional entropies; any error in the directed rounding or the rational change of basis would remove the rigorous gap even if the floating-point numbers look positive.

What would settle it

Recompute the two entropies of the cutoff output states for the explicit six-level witness at η=0.8, ν=1 with independent interval arithmetic or exact rational arithmetic; if the certified lower endpoint is non-positive, the separation claim fails.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

Share X Bluesky LinkedIn Reddit HN

If this is right

  • Quantum capacity of the thermal attenuator is positive for all η≥0.7841 when ν=1, and on a whole two-dimensional high-ground region of the (η,ν) plane obtained by Gaussian pre- and post-processing.
  • Single-mode Gaussian inputs are provably suboptimal for the one-shot coherent information of the thermal attenuator.
  • Reliable bosonic quantum communication is possible in high-noise regimes previously left open between the antidegradability threshold and the thermal threshold.
  • Any complete capacity theory for bosonic loss-plus-noise channels must include non-Gaussian or multi-mode input optimizations.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same residue-class Fock constructions used for the witnesses may systematically improve coherent-information lower bounds for other phase-insensitive Gaussian channels once the Gaussian optimum is known.
  • Because the numerical gap at the thermal threshold already reaches 8.4×10^{-3}, modest further optimization could produce rates large enough to matter for finite-blocklength bosonic coding.
  • The persistent separation from the antidegradability line suggests a genuine intermediate regime whose exact capacity remains open and may require multi-letter or adaptive strategies.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

0 major / 4 minor

Summary. The paper studies the one-shot coherent information of the bosonic thermal attenuator Φ_{η,ν}. It first proves that the long-standing Holevo–Werner thermal-state lower bound is exactly the supremum of I_c over all single-mode Gaussian inputs (Lemma 1 / SM), via a concavity argument that fixes the output symplectic eigenvalue and compares the pure and thermal endpoints. It then exhibits an explicit rank-two non-Gaussian state ρ★ supported on six Fock levels for which, at the thermal threshold (η,ν)=(4/5,1), a cutoff-erasure lower bound I_c(ρ★,T_11∘Φ) is rigorously certified positive by interval arithmetic and Gershgorin enclosures (Theorem 1, Appendix A), yielding Q(Φ_{4/5,1})≥4.7×10^{-4}>0. A second certified witness extends positivity down to η=0.7841 at ν=1, and data-processing enlarges this to a two-dimensional region of the (η,ν) plane beyond the thermal threshold.

Significance. The result settles two open questions left by Holevo–Werner (1999): Gaussian inputs are optimal among single-mode Gaussians, yet non-Gaussian inputs can strictly outperform them and establish positive quantum capacity in the intermediate window η_AD(ν)<η≤η_G(ν) where the thermal bound vanishes and antidegradability no longer forces Q=0. The analytic Gaussian-optimality proof is clean; the constructive certificates (explicit states, cutoff monotonicity, and documented interval/Gershgorin method) supply concrete, falsifiable lower bounds and identify new high-noise regimes for bosonic quantum communication. Strengths include the transparent SM concavity argument and the independent second certificate at η=0.7841.

minor comments (4)
  1. [Appendix A] Appendix A describes the interval-arithmetic/Gershgorin certificate in careful detail but does not ship code or a machine-checked artifact. Providing a short reproducible script (or a pointer to one) would eliminate the only residual reproducibility caveat.
  2. [Fig. 1] Figure 1(a) logarithmic scale and the open circle at η=0.7818 are helpful; a brief remark in the caption clarifying that the open circle is uncertified would avoid any ambiguity for readers skimming the figure.
  3. [AI-ASSISTED RESEARCH STATEMENT] The AI-assisted research statement is unusually candid and welcome; a one-sentence note on how the residue-class families were subsequently verified independently of the AI suggestion would further strengthen reproducibility optics.
  4. [Supplemental Material] In the SM, the representation (S21)–(S22) of the entropy-exchange function is elegant; a short remark that the same concavity holds for the energy-constrained problem only after a different argument (or does not) would prevent mis-citation.

Circularity Check

0 steps flagged

No significant circularity: Gaussian optimality is a self-contained concavity proof; non-Gaussian positivity is an explicit constructive certificate, not a fitted or self-defined quantity.

full rationale

The paper’s two load-bearing claims do not reduce to their inputs by construction. Lemma 1 equates the single-mode Gaussian coherent-information supremum to the Holevo–Werner/Brádler thermal value via a new fixed-output-entropy concavity argument on the entropy exchange (SM Eqs. S17–S28); thermal optimality is an independent prior result, not redefined here. Theorem 1 and Eq. (22) are explicit lower bounds: a concrete finite-support state is fed through a cutoff-erasure channel, and positivity of I_c is certified by interval arithmetic on Kraus matrix elements (Appendix A). That is a one-sided constructive witness, not a fit renamed as a prediction, and the target (Gaussian optimum = 0 at η_G) is established separately. Self-citations (e.g. high-ground region H_{η0,ν0} from Kianvash–Fanizza–Giovannetti) only enlarge the positivity region by data processing after the certificate is already in hand; they are not used to force the central separation. No self-definitional loop, fitted-input-as-prediction, or uniqueness-import circularity appears in the derivation chain.

Axiom & Free-Parameter Ledger

3 free parameters · 6 axioms · 2 invented entities

The work sits on standard continuous-variable QI (LSD theorem, coherent information, Gaussian state calculus, known degradability/antidegradability thresholds) plus two external analytic results (Holevo–Werner thermal coherent information and Brádler’s infinite-energy supremum). The only paper-specific ingredients are the hand/AI-chosen non-Gaussian Fock coefficients and the computer-assisted interval certificate; no new physical entities are postulated.

free parameters (3)
  • Fock coefficients of |e⟩, |o⟩ in ρ★ = ρ★ = (1/3)|e⟩⟨e| + (2/3)|o⟩⟨o| with |e⟩∝5|0⟩-4|4⟩+2|8⟩, |o⟩∝6|1⟩-5|5⟩+2|9⟩
    Integer coefficients (5,-4,2) and (6,-5,2) and mixture weights 1/3, 2/3 are chosen by search so that certified I_c > 0 at η=4/5, ν=1; they are constructive witnesses, not fits to external experimental data.
  • Fock coefficients of |e0⟩, |o0⟩ in ρ0 = weights 67/182 and 115/182; a_j, b_j as listed in Appendix B
    Ten-term modulo-4 coefficient vectors and weights 67/182, 115/182 chosen to certify positivity at η=0.7841; again constructive, not data-fitted.
  • Cutoff M in T_M = M=11 and M=50
    M=11 (Theorem 1) and M=50 (Eq. 22) are chosen large enough for a positive certified gap; they affect only the lower bound tightness, not the channel model.
axioms (6)
  • domain assumption Lloyd–Shor–Devetak theorem: Q(N) equals the regularized coherent information, so a single-letter I_c > 0 implies Q > 0.
    Invoked to pass from certified I_c(ρ★) > 0 to Q(Φ) ≥ R★ (end of Theorem 1 proof).
  • domain assumption Holevo–Werner evaluation of I_c for thermal inputs and its infinite-energy limit; Brádler’s result that the thermal supremum is attained at infinite energy for η > 1/2.
    Used in Lemma 1 / SM step 3 to identify the Gaussian optimum with [log2(η/(1-η)) - g(ν)]_+.
  • domain assumption Thermal attenuator is antidegradable (hence Q=0) for η ≤ η_AD(ν) = (ν+1/2)/(ν+1).
    Cited from Lami et al. and degradability literature to mark the hard zero-capacity boundary.
  • standard math Data-processing inequality for coherent information: processing the output through T_M can only decrease I_c, so I_c(ρ, Φ) ≥ I_c(ρ, T_M ∘ Φ).
    Step 1 of Theorem 1 proof; standard conditional-entropy monotonicity.
  • standard math Gershgorin’s circle theorem plus outward-rounded interval arithmetic correctly enclose eigenvalues and von Neumann entropies of the explicit finite density matrices.
    Appendix A certification pipeline; standard numerical analysis tools applied to Kraus-built matrices.
  • domain assumption Single-mode Gaussian states are characterized by covariance matrices V_A = t S_r^2 with t ≥ 1; entropy formulas s(x)=g((x-1)/2) and two-mode symplectic eigenvalue identities hold.
    SM Gaussian optimization setup; standard CV Gaussian formalism.
invented entities (2)
  • Cutoff-erasure channel T_M independent evidence
    purpose: Finite-dimensional completely-positive trace-preserving map that retains Fock levels 0…M and flags the rest, enabling rigorous entropy certification while lower-bounding true I_c.
    Defined in Eq. (12) as a proof tool; not a new physical noise model, and independent evidence is unnecessary because it is an explicitly constructed CPTP map.
  • Residue-class non-Gaussian ansatz families (modulo-4 rank-two, modulo-5 rank-three) independent evidence
    purpose: Structured search space for non-Gaussian inputs that yield positive coherent information below the thermal threshold.
    Introduced as numerical/constructive families; the certified witnesses are concrete density operators, not new ontological objects.

pith-pipeline@v1.2.0-daily-grok45 · 20445 in / 3883 out tokens · 84827 ms · 2026-07-31T00:32:14.999873+00:00 · methodology

0 comments
Cite this review

Pith. "Pith review of Bosonic quantum communication beyond the thermal threshold." pith.science (2026). https://pith.science/paper/J6CBP2Q2

@misc{pith2026260727449,
  author       = {Pith},
  title        = {Pith review of: Bosonic quantum communication beyond the thermal threshold},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/J6CBP2Q2}},
  note         = {Machine review of arXiv:2607.27449}
}
Share X Bluesky LinkedIn Reddit HN
read the original abstract

The quantum capacity of the bosonic thermal attenuator, which is given by the regularization of its coherent information, is unknown. The seminal work of Holevo and Werner established in 1999 the standard one-use lower bound obtained from input thermal states. We first prove that this long-standing lower bound is the exact supremum over all single-mode Gaussian states and then show that, crucially, a non-Gaussian state can do better. As a consequence, we prove positivity of the quantum capacity in a parameter region where the channel is not antidegradable, yet its coherent information optimized over single-mode Gaussian states vanishes. For example, with one thermal photon in the environment and at transmissivity $\eta=0.8$, the coherent information is non-positive for every single-mode Gaussian input. We give an explicit rank-two non-Gaussian state, supported on only six Fock levels, whose coherent information is certified to be at least $4.7\times10^{-4}$ qubits per channel use. This short witness is far from numerically optimal: a numerical optimization over fixed non-Gaussian families reaches at least $8.4\times 10^{-3}$ qubits per channel use at the same point. More generally, at $\nu=1$, using non-Gaussian inputs we certify positivity of the coherent information, and therefore of the quantum capacity, down to $\eta=0.7841$; by contrast, the channel is antidegradable, and hence has zero quantum capacity, for $\eta\leq0.75$. Overall, our work identifies new high-noise regimes in which bosonic quantum communication is possible.

Figures

Figures reproduced from arXiv: 2607.27449 by Francesco Anna Mele, Giuseppe Catalano, Ludovico Lami, Marco Fanizza, Vittorio Giovannetti.

Figure 1
Figure 1. Figure 1: FIG. 1. Lower bounds on the quantum capacity of the thermal attenuator [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2. Parameter region [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

68 extracted references · 4 linked inside Pith

  1. [1]

    (12)iscom- pletely positive and trace preserving; for example, it has Kraus operators𝐾0=𝑃 𝑀 and𝐾𝑛=|⊥⟩⟨𝑛| for𝑛> 𝑀

    Thecutoffgivesalowerbound.ThemapinEq. (12)iscom- pletely positive and trace preserving; for example, it has Kraus operators𝐾0=𝑃 𝑀 and𝐾𝑛=|⊥⟩⟨𝑛| for𝑛> 𝑀. Forapurifica- tion𝜓𝑅𝐴 of𝜌𝐴, the coherent information is the negative condi- tional entropy, i.e.,𝐼c(𝜌,N)=−𝑆(𝑅|𝐵) (id⊗N)(𝜓) . Processing 𝐵 through another channel can only increase𝑆(𝑅|𝐵) [1, 2], and hence c...

  2. [2]

    [17, 19, 33])

    The finite matrices are explicit.Every thermal attenuator admits the decomposition Φ𝜂,𝜈 =A 𝐺◦L𝜏, 𝐺=1+(1−𝜂)𝜈, 𝜏=𝜂/𝐺,(15) whereL𝜏 is a pure-loss channel andA𝐺 is a quantum-limited amplifier (see e.g. [17, 19, 33]). At the point in Eq.(8), one has𝐺=6/5 and 𝜏=2/3 . A convenient Fock-basis Kraus representation for the two quantum-limited factors is [33, 58] 𝐿ℓ...

  3. [3]

    Evaluation of the finite-dimensional entropies.Let |Ψ⟩𝑅𝐴 = 1√ 3 |0⟩𝑅|𝑒⟩𝐴+ √︂ 2 3|1⟩𝑅|𝑜⟩𝐴,(19) and define 𝜎𝐵′ =(T 11◦Φ 4/5,1)(𝜌★), 𝜎𝑅𝐵′ = id𝑅⊗(T 11◦Φ 4/5,1) (|Ψ⟩⟨Ψ|).(20) The cutoff output has dimension 13, i.e., the Fock levels 0,...,11 together with the erasure flag, so𝜎𝐵′ and𝜎𝑅𝐵′ are, respectively, 13×13 and 26×26 density matrices. A direct numerical ca...

    1938

  4. [4]

    M.M.Wilde,QuantumInformationTheory,2nded.(Cambridge University Press, Cambridge, 2017)

    2017

  5. [5]

    Khatri and M

    S. Khatri and M. M. Wilde, Principles of Quantum Communica- tion Theory: A Modern Approach (2020), arXiv:2011.04672

    Pith/arXiv arXiv 2020

  6. [6]

    Lloyd, Capacity of the noisy quantum channel, Phys

    S. Lloyd, Capacity of the noisy quantum channel, Phys. Rev. A 55, 1613 (1997)

    1997

  7. [7]

    I.DevetakandP.W.Shor,Thecapacityofaquantumchannelfor simultaneoustransmission ofclassical andquantum information, Commun. Math. Phys.256, 287 (2005)

    2005

  8. [8]

    Devetak, The private classical capacity and quantum capacity of a quantum channel, IEEE Trans

    I. Devetak, The private classical capacity and quantum capacity of a quantum channel, IEEE Trans. Inf. Theory51, 44 (2005)

    2005

  9. [9]

    Serafini,Quantum Continuous Variables: A Primer of The- oretical Methods(CRC Press, Taylor & Francis Group, Boca Raton, FL, 2017)

    A. Serafini,Quantum Continuous Variables: A Primer of The- oretical Methods(CRC Press, Taylor & Francis Group, Boca Raton, FL, 2017)

    2017

  10. [10]

    Weedbrook, S

    C. Weedbrook, S. Pirandola, R. García-Patrón, N. J. Cerf, T. C. Ralph, J. H. Shapiro, and S. Lloyd, Gaussian quantum informa- tion, Rev. Mod. Phys.84, 621 (2012)

    2012

  11. [11]

    A. S. Holevo and R. F. Werner, Evaluating capacities of bosonic Gaussian channels, Phys. Rev. A63, 032312 (2001)

    2001

  12. [12]

    M.M.Wolf,D.Pérez-García,andG.Giedke,Quantumcapacities of bosonic channels, Phys. Rev. Lett.98, 130501 (2007)

    2007

  13. [13]

    F.CarusoandV.Giovannetti,DegradabilityofbosonicGaussian channels, Phys. Rev. A74, 062307 (2006)

    2006

  14. [14]

    Caruso, V

    F. Caruso, V. Giovannetti, and A. S. Holevo, One-mode bosonic Gaussianchannels: Afullweak-degradabilityclassification,New J. Phys.8, 310 (2006)

    2006

  15. [15]

    A. S. Holevo, One-mode quantum Gaussian channels: Structure and quantum capacity, Probl. Inf. Transm.43, 1 (2007)

    2007

  16. [16]

    Phys.10, 083030 (2008)

    F.Caruso,J.Eisert,V.Giovannetti,andA.S.Holevo,Multi-mode bosonic Gaussian channels, New J. Phys.10, 083030 (2008)

    2008

  17. [17]

    Brádler, Coherent information of one-mode Gaussian channels—the general case of non-zero added classical noise, J

    K. Brádler, Coherent information of one-mode Gaussian channels—the general case of non-zero added classical noise, J. Phys. A48, 125301 (2015)

    2015

  18. [18]

    Phys.20, 063025 (2018)

    K.Sharma,M.M.Wilde,S.Adhikari,andM.Takeoka,Bounding the energy-constrained quantum and private capacities of phase- insensitive bosonic Gaussian channels, New J. Phys.20, 063025 (2018)

    2018

  19. [19]

    M.M.WildeandH.Qi,Energy-constrainedprivateandquantum capacities of quantum channels, IEEE Trans. Inf. Theory64, 7802 (2018)

    2018

  20. [20]

    Rosati, A

    M. Rosati, A. Mari, and V. Giovannetti, Narrow bounds for the quantum capacity of thermal attenuators, Nat. Commun.9, 4339 (2018)

    2018

  21. [21]

    Y. Lim, S. Lee, J. Kim, and K. Jeong, Upper bounds on the quantum capacity for a general attenuator and amplifier, Phys. Rev. A99, 052326 (2019)

    2019

  22. [22]

    K. Noh, V. V. Albert, and L. Jiang, Quantum capacity bounds of Gaussian thermal loss channels and achievable rates with Gottesman–Kitaev–Preskill codes, IEEE Trans. Inf. Theory65, 2563 (2019)

    2019

  23. [23]

    Commun.11, 457 (2020)

    K.Noh,S.Pirandola,andL.Jiang,Enhancedenergy-constrained quantum communication over bosonic Gaussian channels, Nat. Commun.11, 457 (2020)

    2020

  24. [24]

    M.Fanizza,F.Kianvash,andV.Giovannetti,Estimatingquantum and private capacities of Gaussian channels via degradable extensions, Phys. Rev. Lett.127, 210501 (2021)

    2021

  25. [25]

    Kianvash, M

    F. Kianvash, M. Fanizza, and V. Giovannetti, Low-ground/High- ground capacity regions analysis for bosonic Gaussian channels, Int. J. Quantum Inf.22, 2440005 (2024)

    2024

  26. [26]

    Pirandola, R

    S. Pirandola, R. Laurenza, C. Ottaviani, and L. Banchi, Fun- damental limits of repeaterless quantum communications, Nat. Commun.8, 15043 (2017)

    2017

  27. [27]

    Pirandola, R

    S. Pirandola, R. García-Patrón, S. L. Braunstein, and S. Lloyd, Direct and reverse secret-key capacities of a quantum channel, Phys. Rev. Lett.102, 050503 (2009). 7

    2009

  28. [28]

    García-Patrón, S

    R. García-Patrón, S. Pirandola, S. Lloyd, and J. H. Shapiro, Reverse coherent information, Phys. Rev. Lett.102, 210501 (2009)

    2009

  29. [29]

    Takeoka, S

    M. Takeoka, S. Guha, and M. M. Wilde, The squashed entangle- ment of a quantum channel, IEEE Trans. Inf. Theory60, 4987 (2014)

    2014

  30. [30]

    Takeoka, S

    M. Takeoka, S. Guha, and M. M. Wilde, Fundamental rate-loss tradeoff for optical quantum key distribution, Nat. Commun.5, 5235 (2014)

    2014

  31. [31]

    M. M. Wilde, M. Tomamichel, and M. Berta, Converse bounds for private communication over quantum channels, IEEE Trans. Inf. Theory63, 1792 (2017)

    2017

  32. [32]

    E.KaurandM.M.Wilde,Upperboundsonsecret-keyagreement over lossy thermal bosonic channels, Phys. Rev. A96, 062318 (2017)

    2017

  33. [33]

    Davis, M

    N. Davis, M. E. Shirokov, and M. M. Wilde, Energy-constrained two-way assisted private and quantum capacities of quantum channels, Phys. Rev. A97, 062310 (2018)

    2018

  34. [34]

    Cope,G.Spedalieri,andL.Banchi,Theoryofchannelsimulation and bounds for private communication, Quantum Sci

    S.Pirandola,S.L.Braunstein,R.Laurenza,C.Ottaviani,T.P.W. Cope,G.Spedalieri,andL.Banchi,Theoryofchannelsimulation and bounds for private communication, Quantum Sci. Technol. 3, 035009 (2018)

    2018

  35. [35]

    Pirandola, Tight bounds for private communication over bosonic Gaussian channels based on teleportation simulation with optimal finite resources, Phys

    R.Laurenza,S.Tserkis,L.Banchi,S.L.Braunstein,T.C.Ralph, and S. Pirandola, Tight bounds for private communication over bosonic Gaussian channels based on teleportation simulation with optimal finite resources, Phys. Rev. A100, 042301 (2019)

    2019

  36. [36]

    F. A. Mele, L. Lami, and V. Giovannetti, Maximum tolerable excessnoiseincontinuous-variablequantumkeydistributionand improved lower bound on two-way capacities, Nat. Photonics19, 329 (2025)

    2025

  37. [37]

    Ortolano, S

    G. Ortolano, S. Pirandola, and L. Banchi, Lower bounding the secret-key capacity of bosonic Gaussian channels via optimal Gaussian measurements (2025), arXiv:2512.15502

    arXiv 2025

  38. [38]

    Smith, J

    G. Smith, J. A. Smolin, and J. Yard, Quantum communication withGaussianchannelsofzeroquantumcapacity,Nat.Photonics 5, 624 (2011)

    2011

  39. [39]

    Y. Lim, R. Takagi, G. Adesso, and S. Lee, Activation and superactivation of single-mode Gaussian quantum channels, Phys. Rev. A99, 032337 (2019)

    2019

  40. [40]

    C. Lupo, V. Giovannetti, and S. Mancini, Capacities of lossy bosonic memory channels, Phys. Rev. Lett.104, 030501 (2010)

    2010

  41. [41]

    L.Lami,M.B.Plenio,V.Giovannetti,andA.S.Holevo,Bosonic quantum communication across arbitrarily high loss channels, Phys. Rev. Lett.125, 110504 (2020)

    2020

  42. [42]

    F. A. Mele, L. Lami, and V. Giovannetti, Restoring quantum communicationefficiencyoverhighlossopticalfibers,Phys.Rev. Lett.129, 180501 (2022)

    2022

  43. [43]

    F. A. Mele, L. Lami, and V. Giovannetti, Quantum optical communication in the presence of strong attenuation noise, Phys. Rev. A106, 042437 (2022)

    2022

  44. [44]

    F. A. Mele, G. De Palma, M. Fanizza, V. Giovannetti, and L. Lami, Optical fibers with memory effects and their quantum communication capacities, IEEE Trans. Inf. Theory70, 8844 (2024)

    2024

  45. [45]

    Pirandola, C

    S. Pirandola, C. Ottaviani, C. S. Jacobsen, G. Spedalieri, S. L. Braunstein, T. Gehring, and U. L. Andersen, Environment- assisted bosonic quantum communications, npj Quantum Inf.7, 77 (2021)

    2021

  46. [46]

    Catalano, M

    G. Catalano, M. Fanizza, F. A. Mele, G. De Palma, and V. Gio- vannetti, Convex combinations of bosonic pure-loss channels (2026), arXiv:2604.26874

    Pith/arXiv arXiv 2026

  47. [47]

    M. M. Wilde, J. M. Renes, and S. Guha, Second-order coding rates for pure-loss bosonic channels, Quantum Inf. Process.15, 1289 (2016)

    2016

  48. [48]

    F.A.Mele,G.Barbarino,V.Giovannetti,andM.Fanizza,Achiev- able rates in non-asymptotic bosonic quantum communication (2025), arXiv:2502.05524

    Pith/arXiv arXiv 2025

  49. [49]

    F. A. Mele, G. Barbarino, V. Giovannetti, and M. Fanizza, Non- asymptotic quantum communication on lossy transmission lines with memory (2025), arXiv:2503.13207

    Pith/arXiv arXiv 2025

  50. [50]

    Arqand, L

    A. Arqand, L. Memarzadeh, and S. Mancini, Quantum capacity of a bosonic dephasing channel, Phys. Rev. A102, 042413 (2020)

    2020

  51. [51]

    P.Leviant,Q.Xu,L.Jiang,andS.Rosenblum,Quantumcapacity and codes for the bosonic loss–dephasing channel, Quantum6, 821 (2022)

    2022

  52. [52]

    Lami and M

    L. Lami and M. M. Wilde, Exact solution for the quantum and private capacities of bosonic dephasing channels, Nat. Photonics 17, 525 (2023)

    2023

  53. [53]

    Zhong, C

    C. Zhong, C. Oh, and L. Jiang, Information transmission with continuous variable quantum erasure channels, Quantum7, 939 (2023)

    2023

  54. [54]

    F. A. Mele, F. Salek, V. Giovannetti, and L. Lami, Quantum communication on the bosonic loss–dephasing channel, Phys. Rev. A110, 012460 (2024)

    2024

  55. [55]

    M.LinandK.Noh,ExploringthequantumcapacityofaGaussian random-displacement channel using Gottesman–Kitaev–Preskill codes and maximum-likelihood decoding, Phys. Rev. A111, 052445 (2025)

    2025

  56. [56]

    Zheng, W

    G. Zheng, W. He, G. Lee, K. Noh, and L. Jiang, Performance and achievable rates of the Gottesman–Kitaev–Preskill code for pure-loss and amplification channels, PRX Quantum6, 030314 (2025)

    2025

  57. [57]

    Giovannetti, S

    V. Giovannetti, S. Guha, S. Lloyd, L. Maccone, J. H. Shapiro, and H. P. Yuen, Classical capacity of the lossy bosonic channel: The exact solution, Phys. Rev. Lett.92, 027902 (2004)

    2004

  58. [58]

    V.Giovannetti,A.S.Holevo,andR.García-Patrón,Asolutionof Gaussian optimizer conjecture for quantum channels, Commun. Math. Phys.334, 1553 (2015)

    2015

  59. [59]

    De Palma, D

    G. De Palma, D. Trevisan, and V. Giovannetti, Gaussian states minimize the output entropy of one-mode quantum Gaussian channels, Phys. Rev. Lett.118, 160503 (2017)

    2017

  60. [60]

    L. Lami, S. Khatri, G. Adesso, and M. M. Wilde, Extendibility of bosonic Gaussian states, Phys. Rev. Lett.123, 050501 (2019)

    2019

  61. [61]

    J. S. Ivan, K. K. Sabapathy, and R. Simon, Operator-sum repre- sentationforbosonicGaussianchannels,Phys.Rev.A84,042311 (2011)

    2011

  62. [62]

    R. E. Moore, R. B. Kearfott, and M. J. Cloud,Introduction to Interval Analysis(SIAM, Philadelphia, 2009)

    2009

  63. [63]

    S. M. Rump, Verification methods: Rigorous results using floating-point arithmetic, Acta Numer.19, 287 (2010)

    2010

  64. [64]

    R. S. Varga,Geršgorin and His Circles(Springer-Verlag, Berlin, 2004)

    2004

  65. [65]

    G. H. Golub and C. F. Van Loan,Matrix Computations, 4th ed. (Johns Hopkins University Press, Baltimore, 2013). 8 SUPPLEMENTAL MATERIAL: GAUSSIAN OPTIMIZATION OF THE COHERENT INFORMATION Here we prove Lemma 1 of the main text. Building on the results of Holevo and Werner [8] and Brádler [14] for input thermal states, our main contribution is to show that t...

    2013

  66. [66]

    A general single-mode Gaussian input.—The thermal attenuator is covariant under phase-space displacements and phase rotations. Consequently, changing the first moments or rotating the input changes the receiver and reference–receiver states only by output unitaries and does not affect the coherent information. Every single-mode Gaussian input can therefor...

  67. [67]

    Equation (S6) implies 𝛽2−(𝜂𝑡+𝑐) 2=2𝜂𝑐𝑡[cosh(2𝑟)−1]≥0.(S13) Define 𝑦 := 𝛽−𝑐 𝜂 .(S14) Because the fixed value of𝛽comes from a physical input with𝑡≥1, Eq

    Fixing the output entropy.—We now fix𝛽, and hence the output entropy𝑠(𝛽). Equation (S6) implies 𝛽2−(𝜂𝑡+𝑐) 2=2𝜂𝑐𝑡[cosh(2𝑟)−1]≥0.(S13) Define 𝑦 := 𝛽−𝑐 𝜂 .(S14) Because the fixed value of𝛽comes from a physical input with𝑡≥1, Eq. (S13) implies𝑦≥1and 1≤𝑡≤𝑦.(S15) Conversely, every𝑡in this interval is compatible with the fixed value of𝛽: choosing𝑟through cosh(2𝑟...

  68. [68]

    Holevo and Werner evaluated their coherent information and its infinite-energy limit [8]

    Optimization of the thermal family.—The remaining optimization over thermal inputs is known. Holevo and Werner evaluated their coherent information and its infinite-energy limit [8]. Brádler subsequently proved that, for1/2< 𝜂 <1, the supremum over the thermal family is either the zero value attained by the vacuum or the value approached in the infinite-e...